SEARCH

SEARCH BY CITATION

Keywords:

  • Drg1;
  • Lerepo4;
  • potassium-dependent stimulation;
  • ribosome binding GTPase;
  • TGS domain

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

Human Drg1, a guanine nucleotide binding protein conserved in archaea and eukaryotes, is regulated by Lerepo4. Together they form a complex which interacts with translating ribosomes. Here we have purified and characterized the GTPase activity of Drg1 and three variants, a shortened mutant depleted of the TGS domain, a phosphomimicking mutant and a construct with the two combined mutations. Our data reveal that potassium strongly stimulates the GTPase activity, without changing the monomeric status of Drg1 and that this activity is notably reduced in the mutants. The nature of Lerepo4 association has also been investigated. Dissecting the role of the different domains revealed that Dfrp domain is the sole responsible for the Drg1 increase in thermal stability and the four fold stimulation over its catalytic activity. Lerepo4 action leaves Drg1 affinity for nucleotides unaffected, feasibly favoring a switch I reorientation, mainly via the TGS domain. Drg1 displayed a high temperature optimum of activity at 42°C, suggesting the ability of being active under possible heat stress conditions.

Structured digital abstract


Abbreviations
BME

β-mercaptoethanol

Dfrp domain

DRG family regulatory protein domain

GAP

GTPase activating protein

GEF

guanine-nucleotide exchange factor

HAS-GTPase

hydrophobic amino acid substituted for catalytic glutamine

TEES superfamily

TrmE-Era-EngA-YihA-Septin-like superfamily

TGS domain

ThrRS, GTPase, SpoT domain

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

Drg1 is a human GTPase belonging to the Obg family that is found in archaea and in eukaryotes but not in eubacteria [1, 2]. It was first discovered as a developmentally regulated GTPase, abundantly transcribed in growing cells in mammalian and frog embryos [1, 3, 4] and in actively growing tissues in plants [5, 6], being predicted to regulate cell growth [5, 7]. It associates with polysomes, playing a role in translation [8].

DRG proteins from nearly all organisms contain 365–370 residues, with several well differentiated domains: the N-terminal HTH domain, the canonical GTP binding domain, the S5D2L insertion domain and a small C-terminal ThrRS, GTPase, SpoT (TGS) domain [9]. The G domain responds to the classical architecture of GTPases with the five G boxes and two switch regions [10], whereas the TGS domain has a predominantly β-sheet structure sharing a common basic fold similar to ubiquitin and sumo proteins (PDB code 2EKI). The TGS domain appears in other members of the family such as YchF and Obg, and also in threonyl tRNA synthetases, suggesting a regulatory role [11].

GTPases are typically very inefficient enzymes; thus many of them become associated with specific regulators, the GDP–GTP exchange factors (GEFs) to mediate their activation, and the GTP-hydrolysis activating proteins (GAPs) to terminate their activity [12]. GAPs usually provide a catalytic residue to the active center that functions in trans, while GEFs help to lower the affinity for the guanine nucleotides. Alternatively, GAP-independent ways of activation have been found such as dimerization, association with ribosomal subunits or cation binding, usually a potassium ion [13]. Neither GAPs nor GEFs have been found for Drg1, which according to its sequence has been predicted to be a potassium-dependent GTPase [13].

Nucleotide cleavage in GTPases is triggered by a water molecule conducting a nucleophilic attack on the gamma phosphate of GTP. There are different ways of positioning or aligning the attacking water molecule and stabilizing the transition state of the phosphoryl transfer reaction [10]. Apart from a catalytic residue generally provided by switch II, the catalytic machinery can be further reinforced by a positive charge introduced by a GAP in trans (usually the so-called arginine finger or asparagine thumb, as in Rho and Ran, respectively [14]). Alternatively, the same role can be taken by the binding of a monovalent cation, such as potassium ion (chemical GAP), coordinated by switch I. The potassium contribution has recently been unraveled thanks to the solved structures of MnmE and FeoB in their active and inactive states [15, 16]. Surprisingly, even though the main catalytic residue is usually found in a conserved position (Gln61 from Ras) in switch II, a novel group of GTPases with a hydrophobic residue instead, called HAS-GTPases [17], has been described in which the catalytic residue either has been replaced by a network of backbone contacts, as in FeoB [15], or is provided by new regions of the protein, such as Glu282 from MnmE [16].

After sequence alignments and the solved structure of the yeast homolog Rbg1 (PDB code 4A9A), DRG factors can be considered as HAS-GTPases, provided that we observe that instead of the catalytic glutamine there is a hydrophobic residue, Ile122, in switch II adopting a ‘retracted conformation’. On the other hand, although there is considerable conformational variability in the switch regions in the absence of a nucleotide, and they can adopt different conformations according to the crystal packing, the arrangement of switch I resembles that of FeoB or MnmE complexed with GDP [15, 16].

Interestingly, although Drg1 is not essential, two and three highly similar paralog proteins can be found in animals and plants, respectively, suggesting an important role of these redundant proteins in fundamental biological processes [1, 8]. In this sense, microarray data and reporter gene assays have shown that DRGs from Arabidopsis thaliana are transcriptionally regulated, and the protein accumulates directly and specifically in response to heat stress and desiccation [18, 19]. Besides, Drg1 from Candida albicans plays a role in the control of invasive filamentation [20]. Lastly, no clear phenotype has been found for the Drg1 or Drg2 knockdown human cell lines [21] or for the yeast gene deletion strains [22]. Taken together, these observations suggest that Drg1 may require certain specific conditions to switch on its regulatory function through the protein synthesis machinery.

In this regard, the biological functions of the members of two subfamilies of the Obg family of GTPases which have been better characterized to date: Obg and YchF have been related to stress response and ribosome assembly [23-28].

Drg1 is subjected to a tight regulation. It has been found to be a target of sumoylation, stimulated by the MEKK1 Map3kinase [29], of ubiquitination [29, 30] and of phosphorylation, undertaken by MPSK1 [31]. The physiological meaning of the latter action is still unclear, but other GTPases are downregulated or upregulated by phosphorylation [32].

Lerepo4 (or Dfrp1) was first identified in vitro and further confirmed in vivo as a partner of Drg1 [1, 8, 30], and it has been demonstrated in yeast to be responsible for its recruitment to polysomes [9, 33]. Automated predictions have considered Lerepo4 as an intrinsically unstructured protein, and the structure of the C-terminal part of Tma46 solved in complex with Rbg1 corroborates this idea as it adopts a non-globular extended conformation, contacting the GTPase and TGS domains of Rbg1 [9].

Lerepo4, highly conserved among eukaryotes, has two unique CCCH-type N-terminal zinc fingers and a C-terminal Dfrp domain, being the second responsible for the interaction with Drg1 [9]. The two consecutive zinc fingers of Lerepo4 are similar to the C3HC4-type and C3H2C3-type RING fingers, characteristic of proteins involved in degradation through the ubiquitin proteasome pathway. Lerepo4 is somehow related to this pathway, since it diminishes Drg1 protein degradation by preventing its polyubiquitination [30], thus representing a way of regulating Drg1 protein levels, which may be important for its function.

Additional roles of Lerepo4 are poorly understood. However, recent data have revealed that human cells increase Lerepo4 expression upon HIV infection [34], and it is also upregulated in neuronal rat cells by neurotrophic factor stimulation [35].

To gain further insights into the function of Drg1, the role of the TGS domain, the mode of interaction with Lerepo4 and its effects on GTPase activity, the kinetic characteristics of Drg1 have been addressed in the present work.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

Characterization of Drg1 enzymatic activity

In order to characterize human Drg1, the full length and the different variants were overexpressed in Escherichia coli and purified nearly to homogeneity (Fig. 1A,B). The kinetic parameters of GTP hydrolysis were analyzed and the best fit of the data was obtained using the Michaelis–Menten equations. Although Drg1 catalyzes the production of equimolar amounts of GDP and Pi at a slow rate, it is faster than the rate observed for other Obg-like GTPases [36, 37]. Kcat is 0.147 min−1 (Vmax = 3.59 ± 0.04 nmol·min−1·mg−1 protein), and Km for GTP is 0.25 mm (Table 1). This enzyme is relatively stable in a wide range of pH values (from 6.5 to 9.0), with an optimal pH extending from 8 to 9 (Fig. S1A). Surprisingly, Drg1 displays a thermophilic behavior. Its activity shows a sharp increase between 37 and 50 °C, almost doubling the activity values for temperatures outside this interval (Fig. S1B). The optimal temperature is consequently relatively high (42 °C, Fig. S1B). Interestingly, Drg1 needs large amounts of potassium ions (Fig. S1C) with the enzymatic activity increasing almost steadily until the highest K+ concentration tested (1 m), whereas it appears to be sodium independent (data not shown). Drg1 transformation of ATP into ADP remained undetected, pointing towards a GTP/GDP specific enzyme (data not illustrated), in contrast to other NTPases from the Obg family, like the YchF subfamily, which are able to hydrolyze both GTP and ATP molecules [38]. This is not surprising, as the guanine specificity is mostly conferred by the G4 motif which in Drg1 fits into the consensus sequence.

Table 1. Kinetic parameters for GTP hydrolysis of Drg1 and the mutant proteins in the presence or absence of the C-terminal part of Lerepo4. Values for GTP, GDP and potassium were obtained from concentration–activity plots fitted to hyperbolae and hyperbolic inhibition kinetics (see 'Characterization of Drg1 enzymatic activity'). Standard deviations are given
 GTPGDPK+
Kcat (min−1)Km (mm)Ki (mm)Km (m)
  1. a

     Apparent value, not reliable. The curves do not reach the steady state, due to insufficient amounts of effector.

Wild-type0.147 ± 0.0020.25 ± 0.010.27 ± 0.031.20 ± 0.13a
T100D0.080 ± 0.0020.26 ± 0.030.23 ± 0.020.31 ± 0.03
Drg1ΔTGS0.090 ± 0.0060.22 ± 0.060.22 ± 0.030.65 ± 0.17
Drg1ΔTGS + T100D0.035 ± 0.0010.19 ± 0.020.29 ± 0.061.70 ± 0.43a
Wild-type + C-terminal part of Lerepo40.580 ± 0.0060.25 ± 0.010.37 ± 0.050.10 ± 0.01
T100D + C-terminal part of Lerepo40.465 ± 0.0060.25 ± 0.010.34 ± 0.020.11 ± 0.01
Drg1ΔTGS + C-terminal part of Lerepo40.181 ± 0.0060.19 ± 0.030.35 ± 0.030.60 ± 0.30
Drg1ΔTGS + T100D + C-terminal part of Lerepo40.070 ± 0.0020.19 ± 0.020.28 ± 0.060.60 ± 0.16
image

Figure 1. (A) Coomassie-stained SDS/PAGE (10% polyacrylamide) of the purified wild-type and mutant forms from human Drg1 and Lerepo4. St, molecular mass protein marker (Nippon Genetics, Tokyo, Japan) with the masses, in kDa, indicated at the side. The numbers stand for Drg1 (1), Drg1 T100D (2), TGS deleted Drg1 mutant (3), double mutant of Drg1 (4), C-terminal part of Lerepo4 (5), N-terminal part of Lerepo4 (6), full-length Lerepo4 (7), TGS domain (8). (B) Schematic representation of the domain organization of the different constructs used in this work. Figures denote residue numbers.

Download figure to PowerPoint

Lerepo4 enhances the activity of Drg1

Despite the fact that Drg1 is able to catalyze GTP hydrolysis without auxiliary factors, the activity can be modulated in several ways. Apart from the mentioned effect of K+, Lerepo4 is the most determining factor on catalysis. The activity of Drg1 is around four times higher in the presence of this partner (Fig. 2A). Kinetic assays demonstrate that the ratio between Lerepo4 and Drg1 is 1 : 1 (Fig. 2B). This is equivalent to the association of the yeast homologs Rbg1 and Tma46, which is a heterodimer made of a monomer of each protein, as has been illustrated by the recently solved crystal structure of the complex [9]. While the biophysical aspects of Drg1, such as the optimum temperature and pH, remain the same, the kinetic constant for potassium improves when Lerepo4 is present (Fig. S1, Table 1).

image

Figure 2. (A) Influence of Lerepo4 full length (●), Lerepo4 C-terminal part (○) and Lerepo4 N-terminal part (*) on GTP saturation curves of Drg1 with respect to the activity of Drg1 alone (□). The assay contained 20 mm MgCl2, 300 mm KCl and increasing concentrations of GTP. (B) Dependence of the activity of Drg1 on molar amounts of Lerepo4 full length (●), Lerepo4 C-terminal part (○) and Lerepo4 N-terminal part (*) with respect to the Drg1 activity alone, considered as the zero value; and that of the TGS truncated variant regarding the C-terminal part of Lerepo4 (▲).

Download figure to PowerPoint

The N-terminal part of Lerepo4 has no effect on Drg1 activity whereas the C- terminal part is responsible for the GTPase stimulation

No increase of activity was observed when the Drg1 GTPase was evaluated in the presence of the N-terminal part of Lerepo4, corresponding to the zinc fingers, even when the molar ratio was increased 10 times with respect to Drg1. However, Lerepo4 Dfrp domain is enough to induce the reported increase in Drg1 activity (Fig. 2A) to the same extent as full-length Lerepo4. The ratio between the C-terminal part of Lerepo4 and Drg1 is also shown (Fig. 2B), indicating a 1 : 1 stoichiometry.

GDP inhibition is competitive

Drg1 is completely and hyperbolically inhibited by GDP when assayed at 1.5 mm GTP, with a Ki of 0.27 mm (Fig. S2A, Table 1). The inhibition constant did not improve in the presence of Lerepo4, showing no real contribution of Lerepo4 in GDP binding. Lineweaver–Burk graphs show that the inhibition was competitive, since its mainly detrimental effect on the activity of the enzyme was an increase in Km (Fig. S2B).

Mimicking phosphorylation of Thr100 decreases Drg1 activity

Drg1 was identified as a substrate of the human protein kinase MPSK1 and was shown to be phosphorylated on Thr100, an evolutionarily conserved residue located immediately after the G2 motif, as part of switch I [31]. To ensure reproducible enzymatic assays, instead of doing in vitro phosphorylation of the wild-type enzyme, phosphorylation was mimicked through a mutation to an aspartate residue that simulates the negative charges present on phosphorylated serines. T100D mutation triggers a drop in the activity, producing an enzyme two-thirds less active, suggesting that phosphorylation could be a means of downregulating the enzyme (Table 1, Fig. S3). On the other hand, the mutation does not have a significant effect on GTP binding or GDP inhibition (Fig. S2B).

TGS domain is contributing to Drg1 activity and binding to Lerepo4

Although the GTPase activity is within the G domain, the TGS domain has a modulatory effect over it, as reflected by a decrease of the maximum velocity of the truncated form (Table 1, Fig. S3). As expected, the TGS domain on its own does not show any activity (data not illustrated). GDP inhibited identically the wild-type and the TGS truncated enzyme form (Fig. S2B), demonstrating that the Drg1 TGS domain is not involved in GDP mediated inhibition. Similarly, GTP binding was unaffected.

Interestingly, the data show that in the absence of Lerepo4 the activity of the TGS truncated form was greater than the activity of the phosphorylation mimicking mutant. However, the opposite holds true in the presence of Lerepo4, as the truncated form lacks a great percentage of the area used to interact with Lerepo4 (Fig. S3). Nevertheless, the TGS truncated Drg1 becomes associated with the C-terminal part of Lerepo4 with a 1 : 1 ratio (Fig. 2B).

The TGS deletion when combined with the phosphorylation mimicking mutation produced an additive effect, making the enzyme less active and susceptible to Lerepo4 stimulation.

Drg1 is thermally stabilized by Lerepo4

The incubation of Drg1 with Lerepo4 at various temperatures prior to the enzymatic measurement of the activity showed a protective effect of Lerepo4 over Drg1 GTPase activity. This effect was very evident at 60 °C, where after 5 min of incubation around 70% of the activity was retained when Drg1 was accompanied by Lerepo4 but only 10% activity was maintained by Drg1 on its own (Fig. 3A).

image

Figure 3. Wild-type Drg1 and the three variants were incubated at the indicated temperatures with or without a molar excess of the C-terminal part of Lerepo4: (A) wild-type Drg1; (B) the phosphomimicking mutant; (C) the TGS truncated mutant; (D) the TGS truncated mutant with the phosphomimicking mutation. Samples were taken after the indicated periods of time for assays of enzyme activity at 37 °C. This activity is represented as a percentage of the activity shown before incubation. Values at 37 °C in the absence of the C-terminal part of Lerepo4 are shown as full circles (●), at 45 °C as full squares (■) and at 60 °C as full triangles (▲), connected by a dotted line, whereas in the presence of the C-terminal part of Lerepo4 at 37 °C they are shown as open circles (○), at 45 °C as open squares (□) and at 60 °C as open triangles (Δ).

Download figure to PowerPoint

The phosphorylation emulating mutant T100D displayed only a slightly decreased effect on the Drg1 GTPase activity pre-incubated on its own (Fig. 3B). The TGS domain deletion, on the other hand, showed a reduction in the Lerepo4 mediated thermal protection that became evident at 60 °C (Fig. 3C).

The combination of the phosphomimicking mutation plus the TGS deletion had a stronger effect over the enzyme, making Drg1 more unstable, noticeably at 37 °C, and less susceptible to Lerepo4 protection, shown as a reduced relative activity at 45 °C (Fig. 3D).

K+-dependent activity is affected by TGS and by phosphorylation

When Drg1 variants were assayed alone, they all showed low affinity for potassium. The truncated mutant showed a slight increase in K+ affinity suggesting a moderate negative influence of TGS domain over the K+ binding conformation of switch I. This effect was more notorious for the phosphomimicking mutant, which in fact introduced a negative charge in place of Thr100, possibly affecting the positioning of the loop in a favorable manner towards K+. However, when both mutations were combined, K+ affinity was similar to Drg1.

In the presence of Lerepo4, the Km for potassium of wild-type Drg1 experienced a 10-fold improvement; the phosphomicking mutant reached the same value, while the truncated versions of the enzyme were unable to improve the binding of potassium substantially (Table 1, Fig. S4).

K+ does not promote the dimerization of Drg1

The apparent state of Drg1 deduced by the elution volume of size exclusion chromatography, 82 mL, with a buffer without potassium is that of a monomer (data not illustrated). The same chromatogram was obtained when the running buffer included K+, or K+ plus GDP plus aluminium trifluoride, denoting that K+ does not play a role in Drg1 dimerization as an additional way of activating the enzyme. This is opposite to other GTPases such as MnmE, a tRNA modification GTPase belonging to the Era-related GTPases [16], but akin to YchF, an ATPase from its family, or to EngA [39, 40]. A similar result was achieved through differential ultracentrifugation techniques. The c(s) profile of the protein obtained in the velocity experiment analysis showed three peaks at 0.8 mg·mL−1, regardless of the buffer used: the predominant component, at 72.2%, had a sedimentation coefficient of 2.6 ± 0.1 S, which is equivalent to a molecular weight of 56 kDa, close to the size of the monomer of Drg1 (Fig. S5). A minor proportion of higher molecular weights with apparent molecular masses close to Drg1 dimer or tetramer were also detected (Fig. S5); however, this was independent of the presence/absence of potassium and/or GDP plus aluminium trifluoride indicating that potassium is not regulating the oligomeric state of the protein.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

For the first time, we have accomplished a detailed biochemical study of a GTPase from the Drg1 subfamily, filling the gap between all the structural and functional knowledge gathered in the past years about DRG factors and the lack of data about its kinetic performance.

No need for GAPs and GEFs

Drg1 exhibits a reasonable intrinsic GTP hydrolysis rate of 0.15 min−1 (greater than the values for Ras proteins, which are ~ 0.02 min−1), suggesting that there is no need for a GAP, in contrast to the small GTP binding proteins of the Ras superfamily [14]. The activity of Drg1 is not much lower than that of other GTPases from the TEES superfamily, such as FeoB or EngA [15, 40]. Likewise, Drg1 has micromolar binding constants for GTP and GDP, which give rise to a rapid exchange of the nucleotide, and hence it does not entail the presence of GEFs.

These properties indicate that in vitro the GTPase cycle of Drg1 occurs efficiently without auxiliary factors. Although the catalytic mechanism remains unknown, our results substantiate the importance of K+, confirming previous bioinformatic predictions, even though there are serine residues instead of the two conserved asparagines in the region of the G1 motif [13]. Furthermore, we have expanded knowledge by demonstrating that sodium fails to stimulate Drg1 activity, which compels us to include it within group I of potassium-selective cation-dependent G proteins [13].

Proposed model of Lerepo4 stimulation of Drg1 activity

In addition, Lerepo4, the partner protein which interacts with Drg1 in translating ribosomes, was demonstrated to be a modulator of the enzyme in several ways. As presented previously, binding of Lerepo4 to Drg1 is essential to maintain normal levels of chicken and human Drg1 in vivo [30] and to prevent its ubiquitin mediated degradation [30]. Here it has been shown that the physical interaction with the Dfrp domain of Lerepo4 protects Drg1, conferring a greater thermal stability. Additionally Lerepo4 also improves Drg1 hydrolysis rate, rendering a more efficient enzyme. YchF experiences a similar effect, such as when interacting with ribosomal proteins with a 10-fold increase in activity [26].

The solved structure of the yeast homologs Rbg1 and Tma46 in complex almost rules out a role of Lerepo4 homolog as a GAP since they bind to their GTPase partners in the GTP-bound state whereas Tma46 binds to Rbg1 in the absence of nucleotide. Moreover, the Rbg1–Tma46 structure shows Tma46 grasping Rbg1 opposite to the GTP binding site, displaying an open accessible area around the active center. Thus, it seems very improbable that Tma46 could provide any residue to the active center (Fig. 4). However, it should be pointed out that binding of Tma46 might condition the Rbg1 switch I orientation, as the C-terminal end of Tma46 contacts the hinge of this loop sideways (Fig. 5).

image

Figure 4. Overall structure of the Rbg1–Tma46 complex. Tma46 interacts mainly with an Rbg1 area opposite to the GTP binding site. From the N to C terminus, Tma46 interacts with the back of the Rbg1 molecule, all around the G domain, and then goes to wrap the TGS domain, to finally go down to the front of the molecule, ending close to the hinge of switch I. Although Rbg1 has been solved in complex with only the Dfrp domain of Tma46, the structure suggests that neither the zinc fingers of Tma46 nor the HTH and S5D2L domains of Rbg1 are likely to be involved in the binding. Rbg1 is depicted in blue, while Tma46 is depicted in green.

Download figure to PowerPoint

image

Figure 5. Superimposition of the GDP–AlFx (red) and GDP bound (orange) switch I structures of FeoB, together with the apo form of Rbg1 (blue). Residues Thr35, Ile57 from FeoB and Val101, Ile122 from Rbg1 are shown as sticks. The C-terminal end of Tma46 is also shown (in green), contacting Gly87 and Glu89 from Rbg1, the hinge from where switch I emerges.

Download figure to PowerPoint

Reaching this point, a question could arise about the specific role of Lerepo4 in the catalytic machinery. Our findings show that Lerepo4 has no effect on nucleotide binding, improving instead to a high extent the affinity for potassium ions, which indicates that the Lerepo4 positive effect on GTPase catalysis is at least partly mediated by the observed increase in K+ affinity.

As shown in Fig. 5, the superposition of FeoB structures in inactive and active forms with the apo form of Rbg1 shows that switch I from Rbg1 superposes close to the GDP inactive state of FeoB and far away from the GTP binding state. This implies that it could also undergo a large structural change upon hydrolysis of GTP, acting as a lid to close the active center, highlighting the idea of a loop with great conformational mobility that is influencing the catalysis to a large extent. In fact, the mechanism of the GTPases from the Obg family has been described as being produced by a conformational change of switches, triggering internal domain rearrangements leading to an overall shape change of the molecule [10].

Regulatory role of the TGS domain

According to expectations, the TGS domain is unnecessary either for activity or for the inhibition of the enzyme but its presence, although worsening potassium binding, favors Drg1 catalysis, probably through a stabilization of a GTP-induced conformational state. In addition, it has been functionally ratified in yeast that the Rbg1 TGS domain is important for the association with Tma46, the Lerepo4 homolog, and as a consequence this domain plays a primary part in the anchoring of Dfrp domains [9].

Although Lerepo4 still binds the truncated Drg1 variant with a 1 : 1 stoichiometry, it exerts a smaller activation over the mutant and a less protective effect during the high temperature pre-incubation assay, suggesting that the increase in the activity of Drg1 mediated by Lerepo4 takes place mainly through the TGS domain. Moreover, in the structure, Tma46 wraps the TGS domain of Rbg1 and then establishes contacts along the G domain. The structure of the Rbg1–Tma46 complex shows a direct connection between Rbg1, TGS and G domains mediated by helix 2 of Tma46, suggesting that TGS could affect the catalytic properties of Rbg1 through Tma46 [9] (Fig. 4).

Phosphorylation affects switch I conformation

Mimicking the phosphorylation of Drg1 in Thr100 produces an important decay in the activity of the enzyme, suggesting that phosphorylation on Thr100 might be a way to downregulate the GTPase activity of Drg1, with a probable repercussion in its function. In this sense, there is a report of a member of the same family, YchF, in which the function of the protein is regulated by phosphorylation, affecting its stress response [25].

Residue Thr100 is part of switch I, and it is located two residues away from the equivalent Thr35 of FeoB, which has been shown to be important for nucleotide hydrolysis [15]. Nonetheless, in the structure of Rbg1, the equivalent residue (Val101) is pointing away from the active center, probably not participating in catalysis (Fig. 5), implying that the aspartate replacement could be distorting the loop conformation instead and explaining the drop in the catalytic activity. Notwithstanding, this effect can be mostly neutralized by the binding of Lerepo4, which means that Lerepo4 is able to mend the conformation of switch I.

The other relevant effect of this mutation concerns the potassium binding. In the absence of Lerepo4, the affinity for this cation is improved, suggesting a loop conformation in which the residues involved in the binding are better positioned. As a result of these observations, we can conclude that potassium binding is necessary but not enough for catalysis. Other residues and molecular mechanisms must be entangled.

Possible involvement in heat stress

While affecting its activity, Lerepo4 has proved not to interfere with the biophysical characteristics of Drg1, such us optimal pH and temperature. Nevertheless, the enzyme exhibits a maximum pH value, particularly basic, between 8 and 9, and a high temperature maximum with a broad range of temperatures in which it is mostly active (from 37 to 50 °C), which could be a hint about the possible involvement of Drg1 in heat stress, as is the case of the Drg1 plant homolog [19] or other NTPases from its family, such as Obg and OLA1 [27, 28].

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

Taken together, our data show that Drg1 is a low efficient GTPase, with abnormal optimal temperature and pH, possibly downregulated by phosphorylation and dependent on potassium ions, although they do not induce dimerization. The kinetic studies establish a double role for Lerepo4 in the stability and the activity of the enzyme, an effect that appears to reside in the Dfrp domain.

Our results glimpse a parallelism between two different kinds of mutation, a mutation inside switch I and the deletion of the TGS domain. Both mutants dampen their activity and improve potassium binding, indicating that they are affecting catalysis through switch I. Lerepo4 interaction gets to rescue the T100D mutant to almost normal values but not the truncated variant, pointing to TGS as a regulatory domain. Lerepo4 could be capable of affecting Drg1 activity indirectly by binding to the TGS domain, transducing the signal to the G domain, and also directly through contacts between its C-terminal part and switch I, as is hinted at by the solved structure from the yeast homologs (Fig. 4).

In short, we have clarified to some extent the catalytic features of Drg1, and its interaction with Lerepo4, foreseeing a paramount involvement of Lerepo4 in the task performed by Drg1, apart from its already proved protection against ubiquitination and its presence together with Drg1 in the ribosomes.

It would be interesting to determine the crystal structure of Drg1 in complex with a GTP analog to validate the experimental results shown about the architecture of the binding site. Efforts in this direction are presently under way in our laboratory.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

Cloning and generation of mutant forms

Lerepo4 was amplified from the Megaman human transcriptome library (Agilent Technologies, Santa Clara, CA, USA) and cloned into a modified version of pET28 vector containing a cleavage site for PreScission protease enzyme (ppET28, kindly donated by R. Campos). Drg1 was cloned into pProExHTA from an IMAGE clone library. The Drg1 gene was truncated by PCR amplification following the QuickChange method from Stratagene (La Jolla, CA, USA) immediately after an in-frame stop codon that replaced the normal codon number 293. The QuickChange method was also followed to mutate residue Thr100 by an aspartate, in the full-length protein or in the variant lacking the TGS domain. The TGS domain of Drg1 (289–367 amino acids) was subcloned into pET24. The N-terminal part of Lerepo4 (1–226 amino acids) and the C-terminal part of Lerepo4 (221–396 amino acids) were subcloned into ppET28 by PCR.

After confirming the presence of the desired constructs by DNA sequencing, we used the plasmids to transform E. coli BL21 (DE3) or BL21-CodonPlus (DE3)-RIPL cells (Agilent Technologies). Figure 1B shows a chart with a schematic representation of the different constructs.

Protein overexpression and purification

Proteins were overexpressed overnight at 20 °C using 1 mm isopropyl β-d-1-thiogalactopyranoside. Cells were harvested by centrifugation and stored at −80 °C until use. Cell pellets were resuspended in Hepes 20 mm pH 7.5, glycerol 5%, NaCl 0.5 m, β-mercaptoethanol (BME) 2 mm and 0.01% Triton X-100 for sonication with a protease inhibitor cocktail tablet (Complete EDTA-free; Roche, Basel, Switzerland). Samples were centrifuged for 35 min at 30 600 g. Supernatants were loaded on 5 mL nickel loaded chelating columns washed with 20 mm Hepes pH 7.5, glycerol 5%, NaCl 0.5 m, BME 2 mm and 20 mm imidazole and eluted on a gradient up to 500 mm imidazole. Protein-containing fractions were pooled and concentrated. Then, a Superdex 200 High Load 16/60 column fitted into an FPLC system (both from GE Healthcare Life Sciences, Little Chalfont, UK) was used at a flow rate of 1 mL·min−1, at 4 °C, monitoring the A280 of the eluate. The solution used for equilibrating and running the column was Hepes 20 mm pH 7.5, glycerol 15%, NaCl 0.5 m and BME 2 mm. The purified wild-type and variant proteins were stored at −80 °C at high protein concentration (over 15 mg·mL−1) with 15% glycerol.

All variants were abundantly expressed in soluble and stable form (10–20% of the soluble protein in crude E. coli extracts) and purified to homogeneity. Protein purity was assessed visually on SDS 10% or 15% gels. Protein concentration was quantified in a spectrophotometer at 595 nm using Bradford reagent (Bio-Rad, Hercules, CA, USA) and bovine serum albumin as standard.

Enzyme activity assays

GTP hydrolysis was measured with a colorimetric assay for determination of Pi release [41]. The purified proteins (2–50 μm) were incubated with increasing concentrations (0–4 mm) of GTP in 50 μL of binding buffer (100 mm Tris/HCl, pH 8, 300 mm KCl, 20 mm MgCl2, 10% glycerol) at 37 °C for 60 min. Reactions were stopped by adding 200 μL of daily prepared malachite green reagent. This reagent contained 2 volumes of 0.0812% malachite green, 2 volumes of bidistilled water, 1 volume of ammonium molybdate (5.72% in 6 m HCl) and 1 volume of 2.32% polyvinyl alcohol. The samples were read within the next 10 min in a Labsystems Multiskan® Plus plate reader using a 690-nm filter. Blanks containing the corresponding nucleotide concentrations in binding buffer plus malachite green reagent were subtracted from each sample. To quantitate the amounts of enzymatically released Pi, the samples were compared with a standard curve which was prepared with dilutions of a 300 μm KH2PO4 solution in binding buffer, over a range from 0 to 7 nmol of inorganic phosphate. To determine values for Vmax and Km, the data were fitted to the Michaelis–Menten equation using nonlinear regression with graphpad prism software. Inhibition curves were fitted to hyperbolic inhibition kinetics according to the expression V0* [S]/(Ki + [S]). All assays were conducted at least three times, also using different batches of protein purifications.

Analytical ultracentrifugation experiments

An Optima XL-I analytical ultracentrifuge (Beckman-Coulter, Brea, CA, USA) was used to perform the analytical ultracentrifugation experiments. The result was monitored using Rayleigh interferometric detection. Experiments were conducted at 20 °C using an AnTi50 eight-hole rotor and Epon-charcoal standard double sector centerpieces (12-mm optical path). Sedimentation velocity experiments were performed at 48 kr.p.m. using 400 μL samples, at 0.4 and 0.8 mg·mL−1, in buffer consisting of 100 mm Hepes pH 7.5, MgCl2 20 mm, with or without 300 mm KCl, or 300 mm KCl plus GDP 1 mm and AlF3 1 mm. Differential sedimentation coefficient distributions (c(s)) were calculated by least-squares boundary modelling of sedimentation velocity data using sedfit software (version 12.44).

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information

We would like to thank Leticia Domínguez for technical help, Silvia Prado for kinetic assistance and Javier Cervera for his helpful insight. This work was supported by Ministerio de Ciencia e Innovación (SAF2009-10667, SAF2012-31405) and Generalitat Valenciana (Prometeo/2012/061), Spain.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information
  • 1
    Li B & Trueb B (2000) DRG represents a family of two closely related GTP-binding proteins. Biochim Biophys Acta 1491, 196204.
  • 2
    Leipe DD, Wolf YI, Koonin EV & Aravind L (2002) Classification and evolution of P-loop GTPases and related ATPases. J Mol Biol 317, 4172.
  • 3
    Kumar S, Tomooka Y & Noda M (1992) Identification of a set of genes with developmentally down-regulated expression in the mouse brain. Biochem Biophys Res Commun 185, 11551161.
  • 4
    Sazuka T, Tomooka Y, Kathju S, Ikawa Y, Noda M & Kumar S (1992) Identification of a developmentally regulated gene in the mouse central nervous system which encodes a novel proline rich protein. Biochim Biophys Acta 1132, 240248.
  • 5
    Devitt ML, Maas KJ & Stafstrom JP (1999) Characterization of DRGs, developmentally regulated GTP-binding proteins, from pea and Arabidopsis. Plant Mol Biol 39, 7582.
  • 6
    Etheridge N, Trusov Y, Verbelen JP & Botella JR (1999) Characterization of ATDRG1, a member of a new class of GTP-binding proteins in plants. Plant Mol Biol 39, 11131126.
  • 7
    Mahajan MA, Park ST & Sun XH (1996) Association of a novel GTP binding protein, DRG, with TAL oncogenic proteins. Oncogene 12, 23432350.
  • 8
    Nelson BJ, Mass KJ, Dekeyser JML & Stafstrom JP (2009) Association of Drg1 and Drg2 with ribosomes from pea, Arabidopsis and yeast. Int J Plant Sci 170, 834844.
  • 9
    Francis SM, Gas ME, Daugeron MC, Bravo J & Seraphin B (2012) Rbg1–Tma46 dimer structure reveals new functional domains and their role in polysome recruitment. Nucleic Acids Res 40, 1110011114.
  • 10
    Wittinghofer A & Vetter IR (2011) Structure–function relationships of the G domain, a canonical switch motif. Annu Rev Biochem 80, 943971.
  • 11
    Wolf YI, Aravind L, Grishin NV & Koonin EV (1999) Evolution of aminoacyl- tRNA synthetases – analysis of unique domain architectures and phylogenetic trees reveals a complex history of horizontal gene transfer events. Genome Res 9, 689710.
  • 12
    Sprang SR (1997) G protein mechanisms: insights from structural analysis. Annu Rev Biochem 66, 639678.
  • 13
    Ash MR, Maher MJ, Mitchell Guss J & Jormakka M (2012) The cation-dependent G-proteins: in a class of their own. FEBS Lett 586, 22182224.
  • 14
    Bos JL, Rehmann H & Wittinghofer A (2007) GEFs and GAPs: critical elements in the control of small G proteins. Cell 129, 865877.
  • 15
    Ash MR, Maher MJ, Guss JM & Jormakka M (2011) The initiation of GTP hydrolysis by the G-domain of FeoB: insights from a transition-state complex structure. PLoS One 6, e23355.
  • 16
    Scrima A & Wittinghofer A (2006) Dimerisation-dependent GTPase reaction of MnmE: how potassium acts as GTPase-activating element. EMBO J 25, 29402951.
  • 17
    Mishra R, Gara SK, Mishra S & Prakash B (2005) Analysis of GTPases carrying hydrophobic amino acid substitutions in lieu of the catalytic glutamine: implications for GTP hydrolysis. Proteins 59, 332338.
  • 18
    O'Connell A, Robin G, Kobe B & Botella JR (2009) Biochemical characterization of Arabidopsis developmentally regulated G-proteins (DRGs). Protein Expr Purif 67, 8895.
  • 19
    Stafstrom JP (2008) Expression patterns of arabidopsis drg genes: promoter-gus fusions, quantitative real-time PCR, and patterns of protein accumulation in response to environmental stresses. Int J Plant Sci 169, 10461056.
  • 20
    Chen X & Kumamoto CA (2006) A conserved G protein (Drg1p) plays a role in regulation of invasive filamentation in Candida albicans. Microbiology 152, 36913700.
  • 21
    Ishikawa K, Akiyama T, Ito K, Semba K & Inoue J (2009) Independent stabilizations of polysomal Drg1/Dfrp1 complex and non-polysomal Drg2/Dfrp2 complex in mammalian cells. Biochem Biophys Res Commun 390, 552556.
  • 22
    Daugeron MC, Prouteau M, Lacroute F & Seraphin B (2011) The highly conserved eukaryotic DRG factors are required for efficient translation in a manner redundant with the putative RNA helicase Slh1. Nucleic Acids Res 39, 22212233.
  • 23
    Buglino J, Shen V, Hakimian P & Lima CD (2002) Structural and biochemical analysis of the Obg GTP binding protein. Structure 10, 15811592.
  • 24
    Michel B (2005) Obg/CtgA, a signaling protein that controls replication, translation, and morphological development? Dev Cell 8, 300301.
  • 25
    Wenk M, Ba Q, Erichsen V, MacInnes K, Wiese H, Warscheid B & Koch HG (2012) A universally conserved ATPase regulates the oxidative stress response in Escherichia coli. J Biol Chem 287, 4358543598.
  • 26
    Becker M, Gzyl KE, Altamirano AM, Vuong A, Urban K & Wieden HJ (2012) The 70S ribosome modulates the ATPase activity of Escherichia coli YchF. RNA Biol 9, 12881301.
  • 27
    Mao RF, Rubio V, Chen H, Bai L, Mansour OC & Shi ZZ (2013) OLA1 protects cells in heat shock by stabilizing HSP70. Cell Death Dis 4, e491.
  • 28
    Tan J, Jakob U & Bardwell JC (2002) Overexpression of two different GTPases rescues a null mutation in a heat-induced rRNA methyltransferase. J Bacteriol 184, 26922698.
  • 29
    Jakobs A, Himstedt F, Funk M, Korn B, Gaestel M & Niedenthal R (2007) Ubc9 fusion-directed SUMOylation identifies constitutive and inducible SUMOylation. Nucleic Acids Res 35, e109.
  • 30
    Ishikawa K, Azuma S, Ikawa S, Semba K & Inoue J (2005) Identification of DRG family regulatory proteins (DFRPs): specific regulation of DRG1 and DRG2. Genes Cells 10, 139150.
  • 31
    Eswaran J, Bernad A, Ligos JM, Guinea B, Debreczeni JE, Sobott F, Parker SA, Najmanovich R, Turk BE & Knapp S (2008) Structure of the human protein kinase MPSK1 reveals an atypical activation loop architecture. Structure 16, 115124.
  • 32
    Unger H & Golen K (2010) Regulation of Rho GTPase activity through phosphorylation events: a brief overview. In The Rho GTPases in Cancer (Golen K, ed.), pp. 155162. Springer Science+Business Media, Newark, DE.
  • 33
    Fleischer TC, Weaver CM, McAfee KJ, Jennings JL & Link AJ (2006) Systematic identification and functional screens of uncharacterized proteins associated with eukaryotic ribosomal complexes. Genes Dev 20, 12941307.
  • 34
    Capalbo G, Muller-Kuller T, Dietrich U, Hoelzer D, Ottmann OG & Scheuring UJ (2010) Inhibition of X4-tropic HIV type 1 replication by knockdown of the cellular protein LEREPO4. AIDS Res Hum Retroviruses 26, 11551161.
  • 35
    Heese K, Nagai Y & Sawada T (2004) Nerve growth factor (NGF) induces mRNA expression of the new transcription factor protein p48ZnF. Exp Mol Med 36, 130134.
  • 36
    Welsh KM, Trach KA, Folger C & Hoch JA (1994) Biochemical characterization of the essential GTP-binding protein Obg of Bacillus subtilis. J Bacteriol 176, 71617168.
  • 37
    Lin B, Covalle KL & Maddock JR (1999) The Caulobacter crescentus CgtA protein displays unusual guanine nucleotide binding and exchange properties. J Bacteriol 181, 58255832.
  • 38
    Koller-Eichhorn R, Marquardt T, Gail R, Wittinghofer A, Kostrewa D, Kutay U & Kambach C (2007) Human OLA1 defines an ATPase subfamily in the Obg family of GTP-binding proteins. J Biol Chem 282, 1992819937.
  • 39
    Tomar SK, Kumar P & Prakash B (2011) Deciphering the catalytic machinery in a universally conserved ribosome binding ATPase YchF. Biochem Biophys Res Commun 408, 459464.
  • 40
    Foucher AE, Reiser JB, Ebel C, Housset D & Jault JM (2012) Potassium acts as a GTPase-activating element on each nucleotide-binding domain of the essential Bacillus subtilis EngA. PLoS One 7, e46795.
  • 41
    Yim L, Martinez-Vicente M, Villarroya M, Aguado C, Knecht E & Armengod ME (2003) The GTPase activity and C-terminal cysteine of the Escherichia coli MnmE protein are essential for its tRNA modifying function. J Biol Chem 278, 2837828387.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Conclusion
  7. Experimental procedures
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
febs12356-sup-0001-FigsS1-S5.zipapplication/ZIP299K

Fig. S1. Effects of pH (A), temperature (B) and potassium ions (C) on Drg1 enzymatic activity.

Fig. S2. (A) Lineweaver–Burk plot of inhibition kinetics of Drg1 for GDP. (B) Dependence of activity on GDP of Drg1 and the three variants with or without the C-terminal part of Lerepo4.

Fig. S3. Influence of the Drg1 mutations on the concentration dependence of the activity for GTP, based on the absence (A) or presence (B) of the C-terminal part of Lerepo4.

Fig. S4. Influence of the Drg1 mutations on the concentration dependence of the activity for potassium ions.

Fig. S5. Sedimentation velocity data for Drg1.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.